Earthshine blog

On moonlit nights, the sky has a very solar spectrum – this is plainly the case when doing spectroscopy and sky spectra are taken for subtraction from the target object. Since the moon is the reflector, one would expect to have a B-V for the sky like the moon — i.e. ~0.9. On moonless nights, sky brightness contributors are airglow and the extra-solar system background light. Since the former has emission lines, its colour may vary a bit. The latter has solar system, Milky Way and extra-Galactic components. They vary greatly with wavelength — see figure 1 of this paper:

http://aas.aanda.org/articles/aas/full/1998/01/ds1449/node6.html

The zodiacal light has a similar colour to the Sun, in the range 0.2 to 2 microns (i.e UV to near IR) (section 8.1 and figure 38).

Stars in the Milky Way yield a colour for the DGL — diffuse galactic light — of about B-V ~ 0.5 (section 11.1, table 40). I have measured it locally at about B-V~0.6 — see Flynn et al 2006:

http://adsabs.harvard.edu/abs/2006MNRAS.372.1149F

table 5. Note that the light is a lot bluer close to the Sun — few tens of parsecs — the value of ~0.6 is integrated right through the disk in a column. However the colour of the galaxy is very hard to measure and there are a wide range of values in the literature (see table 1 of the Flynn et al 2006 paper).

The dimmest component of the (dark — i.e. moonless) night sky is the EBL – extra galactic background light — which is the radiation field due to distant unresolved galaxies. In the optical this is dominated by light produced in upper main sequence stars and giants. It’s too dim for us to worry about!